A project is proposed to obtain systematic evaluation of the course of recovery from aphasia in a population of stroke patients. The study has three specific goals: First, this project will provide extensive information on the demographic, neuroanatomical, medical and neurolinguistic correlates of the recovery of specific language functions in aphasia. This information about prognostic factors can be used as a data base for the development of on-liner, computer assisted decision aids that would be of use to the neurologist in deciding questions of patient management. Second, the proposed study will evaluate the hypothesis that some language functions recover better than others. Experimental tests will be administered that allow relatively selective evaluation of distinct aspects of language comprehension (such as phoneme discrimination) and of speech production (such as syntactic complexity).. Scores obtained on these measures will be used to evaluate the possibility that there are different recovery rates for particulaar aspects of gross language functions such as comprehension and production. In addition to their considerable theoretical importance, the results of such an evaluation would have significant implications for the design of therapies and communication aids for the aphasic patient. Third, the study proposed here will furnish data for testing hypothesis concerning the functional components that underlie the major aphasic syndromes. Specific issues to be addressed include the incidence of linguistically-defined symptoms (e.g., agrammatism) within the classical syndromes (e.g., Broca's aphasia), and the extent to which the phenomenon of evolution of syndromes during recovery reflects substantive changes in language capacities. This third goal reflects an attempt to join the theories and methods developed in recent neurolinguistic studies of language impairment with the more traditional approach to the study of recovery from aphasia.